The PIRATE PROJECT: a Point-of-care, Informatics-based Randomized, controlled trial for decreasing over-utilization of Antibiotic ThErapy in Gram-negative Bacteremia Dr. Angela Huttner, FMH Division of Infectious Diseases Geneva University Hospitals 5 December 2017 angela.huttner@hcuge.ch
Outline Finding the evidence to support a reduction in antimicrobial usage Point-of-care (POC) randomization trials Learning healthcare systems The PIRATE project Antibiotic resistance & what we should learn from our patients Randomization at the point of care for determining optimal antibiotic durations for Gram-negative bacteremia Substudies
When the drugs don t work because we overused them because we lacked evidence to show that less usage is OK
Hierarchy of evidence Point-of-care randomization studies We don t have enough randomized controlled trials in infectious diseases (only 16% of IDSA recommendations based on them) And even randomized controlled trials may lack external validity Engel et al. Nat Rev Drug Discov 2002; 1(3): 229-37). Spontaneous randomizations occur daily in the clinic, but this evidence goes uncollected (anecdotes) Khan et al. Clin Infect Dis 2010; 51(10): 1147-56
Point-of-care randomization studies Use the electronic health record (EHR) to structure spontaneous pseudo-randomizations at the point of care Enable the coherent study of patient outcomes Data from real patients Follow-up visits integrated into usual care Clinical evidence can come only from the clinic Only suitable for comparing approved treatments or diagnostic techniques toward which there is clinical equipoise Fiore et al. Clin Trials 2011;8(2):183-95
Learning healthcare systems Institute of Medicine (National Academy of Sciences), 2007 : A learning healthcare system is designed to generate and apply the best evidence for the collaborative healthcare choices of each patient and provider; to drive the process of discovery as a natural outgrowth of patient care; and to ensure innovation, quality, safety, and value in health care. https://www.nap.edu/
Common Purpose principles of learning healthcare systems 1. Respect the rights and dignity of patients 2. Respect the clinical judgments of clinicians 3. Provide optimal care to each patient 4. Avoid imposing nonclinical risks and burdens on patients 5. Reduce health inequalities among populations 6. Conduct activities that foster learning from clinical care and clinical information 7. Contribute to the common purpose of improving the quality and value of clinical care and health care systems Faden et al. N Engl J Med 2014; 370(8): 766-8.
Point-of-care randomization How the randomization actually works Minimal electronic health record involvement: Doctor receives a small alert from the EHR that patient could be eligible for the trial Strong EHR involvement: With study team approval, the electronic health record performs the randomization
Establishing a point-of-care randomization platform in Switzerland Huttner A, Leibovici L et al. Closing the evidence gap in infectious disease: point-of-care randomization and informed consent. Clin MicrobioI Inf 2016
Establishing a point-of-care randomization platform in Switzerland Help from above A convincing & easy first test case, with plenty of safety valves Study question of the platform s prototype trial: Are shorter antibiotic courses non-inferior to 14 day courses for Gram-negative bacteremia?
Rationale for the PIRATE project We know we overuse antibiotics We know that this overuse leaves patients with resistant organisms Antibiotic durations are arbitrary and lunar! But physicians are generally afraid to shorten durations without solid (randomized) evidence Holmes AH et al. Lancet. 2015; Bell BG et al. BMC Infect Dis. 2014;14:13; Zarkotou O et al. Clin Microbiol Infect. 2011;17(12):1798-1803.
Are shorter antibiotic courses non-inferior to 14 day courses for Gram-negative bacteremia? Gram-negative bacteremia is on the rise Patients are getting older, more co-morbid, and more immunosuppressed No RCT evaluating the optimal duration of therapy for Gram-negative bacteremia (GNB) has been published Some physicians give 14 days of antibiotics, some 7 and some even only 5 ( pseudo-randomizations ) Indirect evidence that, in patients without structural complications who are improving, shorter durations are safe Havey TC, Fowler RA, Daneman N. Crit Care. 2011;15(6):R267; Sandberg T, Skoog G, Hermansson AB, et al. Lancet. 2012;380(9840):484-490; Chastre J, Wolff M, Fagon JY, et al. JAMA. 2003;290(19):2588-2598
Rationale for the PIRATE project So why not structure these pseudorandomizations at the point of care and follow our patients clinical outcomes?
The PIRATE trial s sites
and team Anne Rossel Laurent Kaiser Stephan Harbarth (The artist) Pierre-Yves Bochud Thomas Perneger Angèle Gayet-Agéron Elodie von Dach Angela Huttner Werner Albrich
Study design Randomized (1:1:1) controlled, triple-blind phase IV POC trial = 500 patients Designated investigators & patients blinded until antibiotic discontinuation Analyst blinded
Population, outcomes Inclusion criteria: 1. Age 18 years 2. The presence of Gramnegative bacteria in at least one blood culture bottle 3. Treatment with a microbiologically efficacious antibiotic Primary outcome = clinical failure = at least one of the following: Exclusion criteria: 1. Immunosuppression 2. Abscess/ other complications 3. Certain difficult organisms 4. Etc. Relapse: a recurrent bacteremia due to the same bacterium occurring from the day of treatment cessation through day 30 Local suppurative complication not present at infection onset Distant complications of the initial infection, defined by growth of the same bacterium causing the initial bacteremia (as determined by antibiotic susceptibility profiling) The restarting of Gram-negative-directed antibiotic therapy due to clinical worsening suspected to be due to the initial infecting organism and for which there is no alternate diagnosis/pathogen suspected Death due to any cause through day 30
Study design: informatics component Electronic-healthcare record workflow for patient identification, randomization and follow-up. The EHR workflow is outlined in red, the control ( back-up ) workflow in grey. Grey arrows indicate safety valves; these cover all points at which the EHR workflow could malfunction. In this hypothetical case, the patient has been randomized to the control arm (antibiotic therapy duration of 14 days).
Study schedule (keeping it simple)
Study timeline Geneva launch 1 April 2017 Lausanne & St. Gallen launch 1 May 2017 Milestones Protocol development Contracts signed Ethics committee approvals Study dissemination, outreach Data management (EHR, ecrf) Patient recruitment Site monitoring Interim analysis Database cleaning, exports Database lock & analyses Manuscript preparation Publication, dissemination Preparation phase Study period Study close-out Year 1 Year 2 Year 3 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Jan - Mar 2017 Apr - Jun Jul - Sept Oct - Dec Jan - Mar 2018 Apr - Jun Jul - Sept Oct - Dec Jan - Mar 2019 Apr - Jun Jul - Sept Oct - Dec (support) X Two-year recruitment period
Opt-out / general information
Ethics Commission s response
Ethics Commission s response
Written vs. oral consent: a substudy!
About that external validity How representative are our included patients? EPCO: Excluded patients outcomes Prospective observational study examining PIRATE s primary outcome in those we didn t enroll
The study that still needs to be done Metagenomic analysis of the microbiota in patients who had shorter versus longer antibiotic courses: Will we indeed decrease the number of antibiotic resistance genes (ARG) that we carry (and transmit)?
Comments, questions? Thank you